Threat Modeling Expert (AI/ML/LLM)

Quick Facts

• Role: Threat Modeling Expert (AI/ML/LLM)

• Engagement: Part-time (~20% of full-time)

• Work model: Remote from within Poland (one-time office visit in Warsaw or Gdańsk to collect equipment)

• Timeline: Start as soon as possible; project duration until end of 2026

• Interview language: English

Role Description

Support a cybersecurity project focused on solutions using AI, Machine Learning, and Large Language Models. Perform application-level security assessments and produce threat models that drive concrete security controls and remediation actions.

Responsibilities

• Conduct application-level security assessments

• Identify assets, trust boundaries, attack surfaces, threats, and security risks

• Map relevant Mythos threat scenarios to application architecture

• Develop and maintain threat models using approved methodologies and tools

• Review solution designs and architecture documentation to validate threat coverage

• Translate identified threats into security controls and remediation actions

• Collaborate with development, architecture, security, and business teams

• Assign owners to mitigation actions and track implementation

• Report identified threats, risks, and remediation progress

Requirements

• Hands-on experience in threat modeling for AI agents, ML, or LLM-based solutions

• Strong practical knowledge of threat modeling methodologies, particularly STRIDE

• Experience with attack trees and/or kill chains

• Ability to analyse assets, trust boundaries, attack surfaces, threat actors, and abuse cases

• Experience in risk assessment and defining appropriate security controls and mitigation measures

• Ability to review application architecture and technical documentation

• Strong communication skills and experience working with business stakeholders

• English proficiency suitable for working in an international environment

Benefits

• Remote work from within Poland with a one-time office visit for equipment pickup

• Part-time engagement (~20% of full-time)

• Project work until end of 2026 with flexible hours arranged individually